PPARα-Target Gene Expression Requires TIS21/BTG2 Gene in Liver of the C57BL/6 Mice under Fasting Condition

Abstract

The TIS21^/BTG2/PC3 gene belongs to the antiproliferative gene (APRO) family and exhibits tumor suppressive activity. However, here we report that TIS21 controls lipid metabolism, rather than cell proliferation, under fasting condition. Using microarray analysis, whole gene expression changes were investigated in liver of TIS21 knockout (TIS21-KO) mice after 20 h fasting and compared with wild type (WT). Peroxisome proliferator-activated receptor alpha (PPARα) target gene expression was almost absent in contrast to increased lipid synthesis in the TIS21-KO mice compared to WT mice. Immunohistochemistry with hematoxylin and eosin staining revealed that lipid deposition was focal in the TIS21-KO liver as opposed to the diffuse and homogeneous pattern in the WT liver after 24 h starvation. In addition, cathepsin E expression was over 10 times higher in the TIS21-KO liver than that in the WT, as opposed to the significant reduction of thioltransferase in both adult and fetal livers. At present, we cannot account for the role of cathepsin E. However, downregulation of glutaredoxin 2 thioltransferase expression might affect hypoxic damage in the TIS21-KO liver. We suggest that the TIS21^/BTG2 gene might be essential to maintain energy metabolism and reducing power in the liver under fasting condition.

Keywords: BTG2; PPARα; fatty acid oxidation; liver metabolism; starvation.

Fig. 1. Starvation-regulated gene expression profile in livers of the TIS21-KO mice analyzed by Agilent gene chips

(A) Total cellular RNAs were isolated from liver of the 10-week old male mice after 20 h starvation and the 13.5 day-old fetus of the C57BL/6 mice with (TIS21-WT) and without (TIS21-KO) TIS21^/BTG2 gene. The mRNAs were purified before analysis with Agilent gene chips. Gene expression profile obtained from the KO mice livers were presented as the relative values based on those of the WT. Venn diagram comparing significantly regulated genes (p < 0.05, ± 2 fold changes). (B) M (log ratio) vs. A (mean average) plot represents each gene as a dot with its relative expression fold (TIS21-KO/TIS21-WT) on Y-axis, and amplitude of the gene (= 1/2 × log₂ [TIS21-WT × TIS21-KO]) on X-axis. Differentially expressed genes (DEGs) selected by the fold are summarized in the Table 1, and the DEGs with different amplitudes were presented in the Table 2. (C) RT-qPCR analysis showing TIS21 expression in liver of C57BL/6 mice normalized by β–actin. (n = 3 per fed group, and n = 6–8 per 24 h fasting group)

Fig. 2. Significant expression of SREBF1/2 target genes, but absence of PPARα target gene induction in the TIS21-KO mice compared to those of the wild type

(A, B) Gene set enrichment analyses showing the increased expression of SREBF1/2 targets, but absence of PPARα target gene expression. (C) Schema showing the biochemical pathways regulated in the TIS21-KO livers based on the DEGs. Note significant increase of Fasn and Scd2 expression as opposed to reduction of the enzymes regulating fatty acid catabolism. Red and green indicate increase and decrease of the gene expression, respectively, in the TIS21-KO liver. (D, E) Validation of the microarray data by RT-qPCR. PPARα target gene expression (D) and SREBF1/2 target gene expression (E) were presented. ^*p < 0.05, ^**p < 0.01 between KO vs. WT. Acad: acyl-CoA dehydrogenase of very long chain (Acadvl), of long chain (Acadl), of medium chain (Acadm), of short chain Acads), of short branched chain (Acadsb); Cpt1a, carnitine palmitoyltransferase 1a; Acox1, acyl-CoA oxidase 1; Sdhb, succinate dehydrogenase b; Fasn, fatty acid synthase; Acaca, acetyl-CoA carboxylase a, Scd, stearoyl-CoA desaturase (1 or 2); Hmgcr, 3-hydroxy-3-methylglutaryl-CoA reductase; Srebf1, steroid response element binding factor 1; Ppar, peroxisome proliferator-activated receptor (alpha or gamma); Ppargc1a, peroxisome proliferator-activated receptor gamma coactivator 1-alpha.

Fig. 3. Histologic changes in the TIS21-KO liver compared with that of the WT in response to starvation

(A) Oil red O staining in male mice liver: Twenty four hour-fasting significantly increased fat contents in liver of both KO and WT mice, compared with those of the fed condition. However, there was no significant difference between the KO and WT. (B) Oil red O staining of female liver: Under fed condition, fat deposition is more in the KO than the WT, whereas 48 h-starvation significantly increased fat contents in both KO and WT mice. (C) Differential effect of TIS21 gene knockout on fat mobilization in liver of male mice after 48 h fasting: Microvesicular fatty change was severe in the WT than that in the KO in response to starvation (compare (c) and (d) panels in Fig 3C), however, centrilobular fat deposition was more severe in the KO than the WT under starvation condition. In fed condition, there was no specific difference between the WT and KO. (D) Differential effect of TIS21 gene knockout on fat mobilization in liver of female mice after 48 h fasting: Microvesicular fatty change was more severe in the WT than that in the KO after starvation. On the other hand, centrilobular focal fat distribution was clear in the KO than the WT under starvation condition. (Compare (c) and (d) panels in Fig. 3D). (E) 24h-fasting significantly increased Triglyceride (TG) and cholesterol level in both WT and KO mice, whereas it was not significantly different between the WT and KO in response to 24h-fasting. Number of the mice used for WT-fed, WT-fasted, KO-fed, KO-fasted are 4,8,2,6, respectively. ^*p < 0.05, ^**p < 0.01, ^***p < 0.0001 between fed and fasted.

Fig. 4. Induction of cathepsin E expression but reduced glutaredoxin 2 in liver of the TIS21-KO mice in both adult and fetus

(A) Heatmap of common DEGs found in the adult and fetal livers of TIS21KO mice. Raw data is displayed in the Supplementary Table S3. (B) RT-qPCR analysis of liver: RNAs were isolated from the 10 week old male mice without fasting and then subjected to the analysis (n = 3 per each group). Note significant increase of cathepsin E (Ctse) as opposed to significant decrease of glutaredoxin 2 (Glrx2) expression in KO livers than those in the WT. Lymphocyte antigen 6 days (Ly6d) expression was also measured, but not significantly different. The data is accordant with that of the cDNA analysis. (C) Cathepsin E expression is increased not only in adult liver but also in other lymphoid organs, such as thymus and splenocytes.